Flying is largely affected by the weather. Before a flight, a pilot needs to study what the weather is likely to be like along the planned route. It is important to account for the wind speed and wind direction so as to know which heading to keep and how long it will take to get to the destination. Perhaps the most important part of the pilot’s pre-planning, however , is to check for weather-related hazards. For larger airplanes, these primarily include turbulence and ice accumulation. For smaller airplanes, especially the ones not allowed to fly through clouds, sight and cloud coverage are fundamental.
These hazards can be encountered in a variety of weather situations. However , there is one type of weather in which they are all present: during active convective weather. This includes thunderstorms, tornados and hurricanes, but also single cell cumulonimbus clouds (CB) without thunder and towering cumulus (TCU), which can cause danger for pilots. Not only do convective events Verbs and the -s endingconjure a broad variety of potential hazards, but they also threaten to produce them with extreme intensity as well as being highly unpredictable.[/annotax]
When the forecasts are wrong concerning convective activity, the consequences can be severe. An example of such an incident occurred on 10 June, 2007, over Stockholm, Sweden. Late in the evening, several warm air balloons were surprised by the development of a convective trough including thunderstorms and heavy rainfall. One balloon had to make a forced landing and an air accident investigation was made by the Swedish Accident Investigation Authority (2007). It was concluded that no forecast or model could have predicted that the trough would develop that evening. Various factors, including the fact that the convective activity commenced from cold air advection at a high altitude, made the development go undiscovered in the model simulations (Swedish Accident Investigation Authority, 2007).
There are a lot of ways to detect convective events, especially if they produce heavy rainfall and/or lightning. However , forecasting them is not as easy. Short range forecasts of a few hours are usually somewhat accurate, but forecasting in the medium and long range is very difficult. Moreover , it is one thing to predict the occurrence of convection over a general area, but quite another to forecast the location, timing, and intensity of single convective cells. One of the major problems with making this kind of forecast is the fact that convective events occur on a very local scale. The size of a developing thunderstorm is for most weather prediction models too small to be resolved at all and assumptions known as parameterizations must be applied. There is also the problem of associated eddies of instability and turbulence being difficult to simulate as a result of very complicated numerical calculations. Another problem is the limited meteorological observations needed as input for the models.
Modellers are constantly trying to overcome these problems. In recent years, much effort has been made to reduce model resolution. A model with resolution of only a few kilometres would for example be able to resolve events of deep convection explicitly without the use of parameterization. However , the area of the model simulations then needs to be quite limited as the computer calculations take a lot more time. Many studies examine the benefits of finer-resolution models and some of them, for example Weusthoff et al. (2010) , look at the way these models handle convective weather specifically.
The purpose of this study is to examine existing studies about the reliability and accuracy of aviation targeted forecasts during events of convection. The literature search is also conducted as a way of investigating the possibility of predicting the specific aviation hazards accompanying these events. The main question to be answered is : With the help of forecast products, for a day with convective activity, what are the odds of a pilot planning a safe flight? This entails examining the reliability of the forecast products available to a pilot before the flight, and the quantity of information gained from them. An additional aspect of the study is to look at the technical details of how a forecast like this is made. What are the important parameters in a model simulation and how good a quality can be produced? The study is limited to products and models in use by Swedish pilots and meteorologists.